Multi-parametric analysis of the cryopreserved bovine semen using imaging flow cytometry with the application of machine learning tools

Abstract

Cryopreservation is an essential technique used in the agricultural sector to preserve cattle semen for a long period of time and further use it for artificial insemination. However, the quality of semen upon thawing decreases drastically, and there are no standardised procedures that could analyse sperm’s morphological and metabolic parameters simultaneously. Based on literature, high variations in methods and results of semen evaluations among studies were determined. Therefore, a meta-analysis was conducted to determine what quality parameters were the main informative ones for a multi-parametric sperm evaluation and to analyse their associations with each other. As results of meta-analysis showed that DNA integrity, mitochondrial membrane potential, and morphology of spermatozoa are important factors for sperm quality evaluation. Advanced technique Imaging Flow Cytometry was used for a rapid, high-throughput, and accurate assessment of each cell’s quality parameters. A novel sperm thawing and staining protocol for multi-parametric evaluation under IFC was developed, and about 100,000 cells were analysed for each quality parameter with the following statistical analysis of obtained quantitative data. This analysis allowed us to see the relationship between mitochondrial membrane potential and abnormal morphology of spermatozoa as well as to characterise a biological mechanism for sperm bundle formation. Image database is currently under creation for the development and application of a comprehensive machine learning algorithm for sperm quality prediction.